FRACTURE BASED APPROACH FOR FAILURE ANALYSIS OF NYLON 600 FIBER - Unika Repository

  

Proceeding of the 11

th

ISSN 114-1284

  Keywords Fracture, failure, nylon 600, pullout

  It is important to recognize the properties of interfaces between fiber and cementitious matrices by a pull-out test and then takes failure

  According to Bazant [16], the failure of concrete structures should consider the strain-softening related to distributed cracking, localized crack that grows to larger fracture prior to failure, and also bridging stresses at the fracture front. Therefore, the suppression of fracture of concrete can be implemented by improving higher toughness and higher tensile ductility [17]. Hence, the application of nylon 600 fiber into cementitious matrix is proper effort to achieve ductility performance.

  A popular approach to fracture problems is fracture mechanics. Fracture mechanics is very important in case of fiber cementitious composites. The improvement of fiber cementitious composites such as FRC, HPRFCC, and ECC seldom implements the fiber application such as nylon, which is categorized as synthetic fiber. It should be noted that fiber takes an important role in determining whole fiber- reinforced cementitious composite (FRC) performance. Previous researches have proved a better performance of ECC using various synthetic fiber surfaces [12], high performance as alike steel performance [13], and even higher compressive stress for irradiated nylon fiber by gamma [14]. The nylon fiber has a special characteristic of multiple constrictions at stretching condition [15] called ‘yield point elongation’ that has magnitude of 200%-300% of initial fiber length.

  ([1]-[11]). Because of its advantages, nylon 600 is believed to be applied into cementitious matrix to achieve best performace of cementitious matrix or concrete.

  Nylon 600 fiber has great value of tension strength, elongation, tension strength, and also elastic modulus ([1], [2], [11]). The embedded nylon 600 fiber provides optimal stable crack length, and improves strain-hardening property

  ) curves of pull-out test will be the same as the load-displacement (P- and stress-strain ( ) curves of fiber tension test.

   International Conference on QiR (Quality in Research) Faculty of Engineering, University of Indonesia, Depok, Indonesia, 3-6 August 2009

  1 , hence do not induce additional fracture; (4) Increasing of strain   after the establishment of stable cracks in point g will increase stress, the second slip will not take place; (5) Broken nylon fibers have a longer embedded length because of the possibility of crack arrester presence is bigger than the shorter ones; and (6) Since the middle right side of matrix is at the intersection point with fiber acts as crack arrester in the beginning of pull-out process, then the load-displacement (P-  and stress-strain (

  1 will not increase stress 

  The research purposes to implement fracture based approach in failure analysis of nylon 600 fiber and also partly embedded nylon 600 fiber in cementitious matrix. The methods are experiment and analytical by modeling. The experiment activities consist of tension test of nylon 600 fiber and pullout test of partly embedded nylon 600 fiber in cementitious matrix with length 150 mm and 180 mm. The research meets conclusions: (1) Whenever fracture takes place, it is always an unstable crack; (2) Stable cracks are established by the presence of crack arrester; (3) After the establishment of stable cracks, increasing strain beyond strain 

  

Tel : (024)844155 hunting Fax : (024)8445265

E-mail : retno_susilorini@yahoo.com; susilorini@unika.ac.id

ABSTRACT

  1 ₁ Department of Civil Engineering

Faculty of Engineering

Soegijapranata Catholic University

  

FRACTURE BASED APPROACH

FOR FAILURE ANALYSIS OF NYLON 600 FIBER

Rr. M.I. Retno Susilorini

1. INTRODUCTION

  Proceeding of the 11 th

ISSN 114-1284

  3. RESULTS AND DISCUSSION

  analysis by fracture based approach. The research purposes to implement fracture based approach in failure analysis of nylon 600 fiber and also partly embedded nylon 600 fiber in cementitious matrix.

   International Conference on QiR (Quality in Research) Faculty of Engineering, University of Indonesia, Depok, Indonesia, 3-6 August 2009

  Figure 2 : Setting for pullout test

2. METHODS

  3.1. Results

  Figure 3 : Relation of stress-strain of nylon 600 tension test The results of pullout test are shown below. Firstly, the experiment results show that the pullout test represent fracture phenomenon in some stages during pull-out process. Figure 4 describes the stages: (a) Pre-slip stage, (b) Slip stage, and (c) Strain-hardening stage. The pre-slip loads are about 400-430 N and pre-slip displacements are no more than 0.1 mm. The slip

  This research conducts experiment and analytical methods. The experiment activities consist of tension test of nylon 600 fiber and also pull-out test of partly embedded nylon 600 fiber in cementitious matrix. The analytical method conveys pull-out modeling.

  70 mm fiber outer length, 30 mm fixed part length, 150 mm partly embedded fiber specimen lower crosshead crank nut upper crosshead nut

  f

  = 250 mm fixed part length, 150 mm embedded length, l

  1

  specimen length, l

  specimen width = 20

  515 517 480 REGRESSION

  The results show that nylon 600 fiber has average maximum tension stress of 1471.21 MPa. Figure 3 shows relation of stress-strain of nylon 600 tension test with average maximum strain of 0.89, average maximum elongation of 84.11%, and average maximum tension load of 1398.13 N. A unique property of nylon fiber is shown by ‘jagged’ phenomenon on stress-strain ( ) curves as also shown by Figure 3. This phenomenon is caused by the yield point elongation and the viscosity of nylon itself.

  1.0 STRAIN S T R E S S (M P a )

  0.8

  0.6

  Figure 1: Dimension of specimen The dimension of specimens is shown by Figure 1 while the setting of pullout test by Figure 3. The pull-out test is provided by computerized Universal Testing Machine “Hung Ta”. The nylon 600 fiber is made in Indonesia with 1.1. mm in diameter and embedded length 150 mm and 180 mm for partly fiber embedded specimens. Mix design for cementitious matrix is cement : sand : water ratio of 1:1:0.6. Analytical method applied by modeling and formulation of theoretical model ([1], [2], [11]). The analytical models are built based on experiment result.

  0.2

  0.0

  1000 1200 1400 1600 1800

  200 400 600 800

  0.4

  th Proceeding of the 11 International Conference on QiR (Quality in Research)

  Faculty of Engineering, University of Indonesia, Depok, Indonesia, 3-6 August 2009

ISSN 114-1284

  loads are in the same range of pre-slip loads with displacements of 3-4 mm. The strain-hardening loads are observed as 1300-1500 N and its displacement about 150-190 mm.

  P P pr P P ps s

  Detail I



pre-slip slip strain-hardening

  Figure 5 : Relation of load-displacement of

   x x 1  2 specimen with embedded length (l ) of 150 mm

  1

  2 f

  and 180 mm

  Detail I

  Figure 4 : The stages during pullout process Figure 5 describes relation of load-displacement of specimen with embedded length (l f ) of 150 mm and 180 mm. Specimen 1025 and 1041 have embedded length l = 150 mm while specimen

  f

  2438 has l = 180 mm. Specimen with l = 180 mm

  f f

  has bigger displacement (190.2 mm) than l = 150

  f

  mm (140.08 and 154.76 mm). On the contrary, specimen with l = 180 mm has lower ultimit load

  f (1400 N) than l = 150 mm (1500-1600 N). f

  Relation of stress-strain of specimen with embedded length (l ) of 150 mm and 180 mm is

  f

  described by Figure 6. Specimen with l = 180 mm

  f

  has bigger strain (2.8) compared to l = 150 mm

  f

  (2.06 and 2.28). On the contrary, specimen with l

  f

  = 180 mm has lower ultimit stress (1500 N) than Figure 6 : Relation of stress-strain of specimen l = 150 mm (1600-1800 N).

  f

  with embedded length (l ) of 150 mm and 180 mm

  f

  The results were analyzed to become basis of pullout modeling. Several aspects have been considered in the modeling: (1) Fracture capacity of embedded fiber is a function of Poisson’s ratio of fiber, (2) Some stages exist during the pull-out and fracture pull-out process, (4) A ‘jagged’ phenomenon exists on strain-hardening part of

  ) and stress-strain () load-displacement (P- curves of pull-out, and (4) Unstable and stable

  Proceeding of the 11 th

   International Conference on QiR (Quality in Research) Faculty of Engineering, University of Indonesia, Depok, Indonesia, 3-6 August 2009

ISSN 114-1284

     

  2 l l (2)

  ratio of total free-end fiber displacement of free-end at stage of strain-hardening

  The range value of E s , E ps , dan E pr for pull-out model is described by Table 1.

  Tabel 1 : Range value of E _s , E _ps , and E _pr STAGE OF STAGE OF SLIP STAGE OF

  PRE- SELIP

  INITIAL SLIP POST SLIP STRAIN- HARDENING l f

  E ps

  E s

  E s

  E pr

  = E n (mm)

  (x 10 ³ MPa) (x 10 ³ MPa) (MPa) (MPa) 150 500 - 650 3000 - 4000 500 -1500 400 - 2500 180 500 - 650 4000 - 5000 500 -1500 400 - 2500

  Pullout modeling also conveys a formulation for stable crack length as expressed by Equation 2.

      

  Where: l

  ratio of total free-end fiber displacement of free-end at stage of slip r

  2 =

  stable crack length (mm) l

  =

  initial outer fiber length (mm)

   =

  fiber displacement at time of stable fracture achieved (mm)

   =

  critical fiber strain Stable crack length for each embedded length can be calculated based on experiment results as explained by Table 2.

  Tabel 2 : Stable Crack Length of Specimens l f l

  2 (mm) (mm)

  150 37,7375 180 37,7375 The pullout model then built based on experiment results and described by Figure 7-10 as follow.

  Figure 6 : Relation of load-displacement of experiment results and model with embedded length l

  f

  = 150 mm

  III =

  II =

     

  load (N) A

       

     

    

     A E a a A r E a a

  A r E a a P r pr

  2

  1 III s

  2

  1 II ps

  2

  1 I n

  (1) Where: P

  n =

  =

  fracture process phenomenon exist during the pull-out process. Pullout modeling conceives a formulation of load which is a function of displacement that is expressed by Equation 1.

  fiber section area (mm

  2

  ) E

  pr =

  modulus of elasticity at stage of strain- hardening (MPa) E

  ps =

  modulus of elasticity at stage of pre-slip (MPa) E

  s =

  modulus of elasticity at stage of slip (MPa) a

  1 =

  total displacement of a stage (mm) a

  2 =

  initial length of specimen or fiber that is specific for every stage (mm) r

  I =

  ratio of total free-end fiber displacement of free-end at stage of pre-slip r

  th Proceeding of the 11 International Conference on QiR (Quality in Research)

  Faculty of Engineering, University of Indonesia, Depok, Indonesia, 3-6 August 2009

ISSN 114-1284

  Figure 8 : Relation of load-displacement of experiment results and model with embedded length l = 180 mm

  f

  Figure 7 : Relation of stress-strain of experiment results and model with embedded length l = 150 mm

  f

  Figure 6 and 7 show that the model fit to the experiment results. For relation of load- Figure 9 : Relation of stress-strain of experiment displacement of experiment results and model results and model with embedded length with embedded length l = 150 mm (Figure 6), the

  f

  l = 180 mm model achieves stage of with load of about 400 N f and stage of strain hardening with load of about 1500 N. Figure 7 shows that for stress-strain

  Figure 8 and 9 also describe that the model fit to relation the model achieved stress at stage of slip experiment results. Figure 8 shows that the model with load about 400 MPa and stage of strain- also achieve stage of slip with load of about 400 hardening with load about 1600 MPa.

  N and stage of strain-hardening with load of about 1350 N. It is lower than the loads achieved by the specimens with embedded length of l f = 150 mm. The same phenomenon happened for stress-strain relation of Figure 9. The model achieve stage of slip with load of about 400 MPa and stage of strain-hardening with load of about 1400 MPa. It is lower than the loads achieved by the specimens with embedded length of l = 150 mm.

  f

  3.2. Discussion

  Whenever fracture takes place, it is always an unstable crack. Unstable crack will change into stable crack with certain condition. Stable crack length will be occured when the crack arrester presents. It means, the phenomenon of unstable fracture that happened at the stage of pre-slip will change into stable fracture at the stage of slip.

  

Proceeding of the 11

th

ISSN 114-1284

  Approach for Structural Element Design – Safe Building, Safe City”, Proceeding Third International Conference on Economic and Urban Management “City Marketing, Heritage, and Identity”, PMLP Unika Soegijapranata, Semarang, pp. 451-465, 2007.

  The author gratefully acknowledges UBCHEA (United Board of Higher Christian Education) for supporting research grant (2005-2007); and to Prof. Ir. Moh. Sahari Besari, MSc., PhD as Promotor; and also to Prof. Bambang Suryoatmono, PhD. as Co-Promotor; for their great contributions of ideas, discussions, and intensive assistance to the dissertation. .

  REFERENCES

  [1] Susilorini, Retno, M.I., Model Masalah Cabut-

  Serat Nylon 600 Tertanam dalam Matriks Sementitis yang Mengalami Fraktur , Dissertation, Unika Parahyangan, Bandung, 2007.

  [2] Susilorini, Retno, Rr. M.I., Pemodelan Cabut-

  Serat Berbasis Fraktur , Unika Soegijapranata Press, 2009.

  [3] Susilorini, Retno, Rr. M.I., “Fractured Based

4. CONCLUSIONS

  Susilorini, Retno, Rr. M.I., “Integral-J Kritis untuk Model Elemen Hingga pada Cabut Serat Fraktur Nylon 600”, Tiga Roda Forum “Perkembangan Terkini Teknologi dan Rekayasa Konstruksi Beton di Indonesia”, 12 Desember, Hotel Bumi Karsa - Bidakara, Jakarta, pp. 1-14, 2007.

  [4]

  e) Broken nylon fibers have a longer embedded length because of the possibility of crack arrester presence is bigger than the shorter ones; f) Since the middle right side of matrix is at the intersection point with fiber acts as crack arrester in the beginning of pull- out process, then the load-displacement (P- ) and stress-strain () curves of pull-out test will be the same as the load- displacement (P- ) and stress-strain ( ) curves of fiber tension test

  [5]

  Susilorini, Retno, Rr. M.I., “Analisa Kegagalan Struktur Berbasis Fraktur untuk Penyelamatan Bumi dan Pembangunan Berkelanjutan”, Makalah, Seminar “Save The Forest, Save The Earth, 5 February, Fakultas Teknik, Unika Soegijapranata, Semarang, pp. 1-11, 2008.

  [6]

  Susilorini, Retno, Rr. M.I., “The Role of Shear- Friction on Pull-Out Fractured Based Modeling of Nylon 600 with Clumped Fiber End”, Seminar Nasional Teknik Sipil IV, 13 Februari, Program Pascasarjana – Jurusan Teknik Sipil, Surabaya, pp. B91-B101.

  [7]

  Susilorini, Retno, Rr. M.I., “Fracture to Failure, a Fracture Mechanics Approach for Bridge Failure Analysis”, Proceeding of International Seminar "The Technology of Long Span Bridge to Strengthen the Unity of Nation", 8-9 July, pp.

  ACKNOWLEDGMENT

  d) Increasing of strain after the establishment of stable cracks in point g will increase stress  (the strain- hardening stage exist), the second slip will not take place;

   International Conference on QiR (Quality in Research) Faculty of Engineering, University of Indonesia, Depok, Indonesia, 3-6 August 2009

  = 150 mm and l

  It is interesting that the load and stress of embeded length l

  f

  = 180 mm are lower compared to l

  f

  = 150 mm. However, the displacement for embeded length l f = 180 mm is bigger compared to l

  f = 150 mm. It can be explained as follow.

  The long embedded length (150 mm and 180 mm) produces broken fiber when specimen gets failure. It should be noted that long embedded length of specimen is related to the possibility of crack arrester presence. Because of that bigger possibility makes the strain-hardening part in load-displacement curve is longer for long embedded fiber length. The stable crack length will be achieved in stable fracture. For both embedded length l

  f

  f

  (the slip stage exists), hence do not induce additional fracture;

  = 180 mm, the stable crack length are the same, l

  2

  = 37,7375 mm (Table 2). It can be seen (Figure 7-10) that they achieve same critical load and stress at the stage of slip that the stable fracture occured. Hence, the modeling represents the fracture phenomenon in appropriate way.

  The research meets conclusions:

  a) Whenever fracture takes place, it is always an unstable crack; b) Stable cracks are established by the presence of crack arrester; c) After the establishment of stable cracks, increasing strain beyond strain 

  1

  will not increase stress 

  1

M.20.1-20.6, 2008.

  th Proceeding of the 11 International Conference on QiR (Quality in Research)

  Faculty of Engineering, University of Indonesia, Depok, Indonesia, 3-6 August 2009

ISSN 114-1284

  [8] Susilorini, Retno, Rr. M.I., “A Fractured Based

  Pull-Out Model of Short Nylon 600 Embedded in Cementitious Matrix”, Jurnal Ilmiah Semesta Teknika (Terakreditasi), Universitas Muhammadiyah Yogyakarta, Vol. 11, No. 1, May, 2008.

  [9] Susilorini, Retno, Rr. M.I., “Stable Crack Length

  on Pull-Out Problem – Significant Factor of Pull- Out Modeling for Concrete Pavement Structure’s Element”, Proceeding of International Symposium

  XI FSTPT, 29-30 October, Universitas Diponegoro, pp. 1-9, 2008.

  [10] Susilorini, Retno, M.I. Rr, “Striving For ‘Green

  Concrete’ with Nylon 600 Fiber - A Review of Pull-Out Model with Nylon 600”, Proceeding of Second Annual International Conference Green Technology and Engineering, 15-17 April, Universitas Malahayati, Lampung, pp. 52-56, 2009.

  Susilorini, Retno, M.I. Rr, “Model Cabut-Serat

  [11]

  Nylon 600 Tertanam dalam Matriks Sementitis Berbasis Fraktur”, Jurnal Dinamika Teknik Sipil (Terakreditasi), published on Vol. 9 No. 1, July Jurusan Teknik Sipil, Universitas Muhammadyah Surakarta, 2009. Li, V.C., Chan, Y.W., Wu, H.C., “Interface

  [12]

  Strengthening Mechanism in Polymeric Fiber Reinforced Cementitious Composites”, Proceedings of International Symposium on Brittle Matrix Composites, (eds. Brandt, A.M, Li, V.C., Marshall, L.H), IKE and Woodhead Publ, Warsaw ,pp. 7-16, 1994.

  [13] Clements, M., “Synthetic as Concrete

  Reinforcement”, Concrete Magazine, United Kingdom, September, pp. 37-38, 2002.

  [14] Martinez-Barrera, G., “Concrete Reinforce with

  Irradiated Nylon Fibers”, Journal of Material Research , Vol.21, No. 2, February, pp. 484-491, 2006.

  [15] Nadai, A., Theory of Flow and Fracture of Solids,

  Volume I, McGraw-Hill Company. Inc, New York, USA, 1950.

  [16] Bazant, ZP., “Fracture Mechanics of Concrete:

  Concepts, Models, and Determination of Material Properties – State of the Art Report”, Proceedings, First International Conference on Facture Mechanics Concrete Structure (Framcos 1), (Ed. Bazant, ZP), Colorado, USA, pp. 6-140, 1992.

  [17] Li, V.C., and Wang, S., “Suppression of Fracture

  Failure of Structures by Composite Design based on Fracture Mechanics”, corresponding paper in Compendium of Papers CD ROM, Paper 5543, 2005.